When the returns from tonight’s election start to become clear, the debate is likely to turn, as it so often does, to why the American people voted the way they did. That turns out to be a very difficult question to answer, however. …
As my Upshot colleague Nate Cohn has noted, exit polls are frequently inaccurate because of their limitations as a survey format. But beyond that, as James Stimson, a political scientist at the University of North Carolina-Chapel Hill, argues in his book “Tides of Consent,” the questions asked in exit polls are rarely useful for understanding why someone voted the way that they did. “What we observe,” he argues, “are mainly post hoc rationalizations.” CONT.
Brendan Nyhan (Dartmouth), New York Times